Cattle grow taller: Implications of outdated ordinal scores for genetic evaluations and selection?

In dairy cattle, conformation and linear-type traits are routinely recorded as categorical or ordinal phenotypes. Some traits originating from continuous biological measurements are discretized into categorical scores. These traits are moderately heritable and play a crucial role in breeding programs, serving as indicators of longevity, fertility, and overall lifetime productivity. However, over time, the underlying phenotypic distributions can evolve due to selection, management, and environmental changes. As a result, the mapping between biological measurements and categorical scores may become obsolete. This paper investigates phenotypic distribution shifts within ordinal scoring systems, using stature in Brown Swiss cattle as a case study. We present an analytical characterization of these shifts and their genetic consequences. A simulation study based on a real pedigree was conducted to quantify the extent of these effects and to explore methods for predicting these changes. The results emphasize periodic recalibration of threshold boundaries to keep the ordinal scoring systems aligned with current population distributions and safeguard the validity of genetic improvement programs. We anticipate that our findings will provide practical guidance for detecting, interpreting, and managing distributional shifts in categorical phenotypes within structured genetic evaluation and selection programs.